Redundant interactions in protein rigid cluster analysis

Folded proteins are held together mainly by weak noncovalent interactions. We propose a method for classifying which of these interactions are critical to maintaining the protein´s 3D shape, using information about their redundancy within the rigid clusters of atoms. We have developed KINARI-Web, a server performing rigidity analysis of proteins. It implements pebble game analysis to determine rigid clusters and flexibility information. The method presented here is provided as an additional module. We classify each of the noncovalent interactions as either redundant or critical. An interaction is critical if removing it would cause the rigid cluster to break apart and become flexible, otherwise an interaction is redundant. In addition, we propose a new method for assigning scores to the rigid clusters based on the number of redundant and critical interactions in the cluster. We present data on the redundancy of the rigid clusters of cytochrome c, barnase, and a dataset of kinases.